[Amps] 1200 vs 1500 triodes

[David A. Pruett]

There is another aspect to this discussion which I think needs to be 
considered.

Most of my operating is during contests, and for that purpose I have a 
large antenna farm with multiple antennas on each band.  The usual amount 
of effort is made to get the best possible match on each antenna.  However, 
the feedline lengths are random and so the actual impedance in the shack 
varies somewhat when the antennas are operated off their zero reactance point.

During contests, many times the operator will switch antennas quickly to 
optimize the signal and make the QSO.  When the load impedance changes, the 
operating point of the tube changes.  One would hope the power output would 
remain nominal, but at the very least the tube should not go into an unsafe 
operating condition.

It has been my experience with the "modern" Eimac triodes (3CX800, 3CX1500) 
that these are rather sensitive to load impedance changes vs. the older 
tubes like 4-1000A, 3-1000Z (of which the 3CX1200A7 is supposedly an 
electrical equivalent).  Let me give some examples.

Example #1: Some years ago my 10M setup consisted of a 6L long boom yagi of 
W1HDQ design, a pair of 4L W2PV yagis in a stack, and a nondescript 4L yagi 
fixed south.  At the time I was using a Ten-Tec Titan (2 x 3CX800) for my 
10M amp and we seemed to have problems when switching antennas so I ran an 
experiment.  I went to a frequency in the 10M band where all three antennas 
showed good SWR (say < 1.2) and all were near their SWR minimum 
point.  After tuneup on antenna #1 the amp put out 1500W with normal grid 
current, on antenna #2 output was 1000W with little grid current, and on 
antenna #3 the output was 500W and the grid meter was on the pin.

Example #2: I was getting ready for the ARRL Sweepstakes and wanted to use 
a homebrew 8877 on 40M.  (The amp was originally built by K8DR, one of the 
well-known Columbus amp builders, although I was not the original 
owner.)  Again, three antennas: a full size 3L yagi, a dipole N/S @ 80' and 
another dipole E/W @ 40'.  Hard as I try, I could not find a single set of 
tune and load settings which would provide safe operation (defined as grid 
current within safe limits) into all three antennas while delivering legal 
limit power.  OTOH, my trusty old Viewstar PT-2500A, running a pair of 
3-500Zs with 3800V no-load, with a single set of tune settings, would 
deliver 1300-1500 to all three antennas with grid current never varying 
more than 20%.

Since these experiences, I've settled on using a single 3-1000Z or 4-1000A 
bottle in my homebrew amps.  Besides the Viewstar, I also have an old Henry 
2K-3 with 3-500Zs, and another K8DR amp using 3-500Zs with 4400V 
no-load.  None of these amps squawk when the antennas are switched.

 From my experiences it seems that the modern tubes with higher gains are 
more sensitive to impedance variation.  When you consider how most of these 
tubes are used in commercial applications, it's no mystery why this is not 
an issue.  The typical commercial amp is connected to a single antenna, 
tuned up on a single frequency and operated that way.  It doesn't see 
impedance variations unless the antenna fails.

I'm not saying that the 3CX800 and 3CX1500 tubes are bad designs.  The 
typical ham who has one antenna on a band will likely be completely happy 
because once tuned up on a band, most frequency excursions are relatively 
small and hence the amp operates somewhat like a commercial 
application.  Also, I have 8877s here on 50 and 144 MHz and operation on 
those bands has been completely satisfactory with multiple antennas, but 
the nature of VHF/UHF being what it is, the antenna impedances tend to be 
held very close to nominal.

While I do have an Alpha 91b with 4CX800A tetrodes, I've not taken the time 
to explore their sensitivity to load impedance changes.

In the end, tube availability may be the most important issue.  I've gotten 
8877s from WA8WZG and highly recommend Tom's stuff.  Some guys use bigger 
bottles like 3CX3000s which are likely to be around for a while.  My 
experiences with the Chinese 3-500Zs has been questionable - a set of RF 
Parts house band tubes shorted out in my Henry 2K-3 (3300V no-load) after 
only 18 months, and this is in an amp which is only used on 80M.  The 
previous Eimacs lasted 20 years...

Like they say, your mileage may vary.

73,

Dave/K8CC



At 08:52 AM 10/24/03 -0600, John T. M. Lyles wrote:
>I vote for the 3CX1500A7/8877-based amplifier also. I helped design an FM 
>broadcast amplifier using it in 1983, and it easily produced 1500 watts of 
>CW output class C at 108 MHz. Last year I converted a RF Power Products 
>plasma generator from 13.56 to ~4-5 MHz for a scientific application. It 
>had a questionable old tube in it, as it was pulled from another 
>experiement in a different lab. When I finished, it fired up immediately 
>at 1200 watts without a burp, no 'gas', no 'oscillations', very stable.
>
>That tube has a myrad of applications, and even though expensive, it is a 
>good part. There have been periods (as with many tubes) where the mfr had 
>a production glitch. However, based on the limited application of the 
>3CX1200 in ham amps, and the domination of commercial and 
>scientific/medical/military apps for the 8877, the answer is fairly obvious.
>John
>K5PRO